Articles to Change how you think about HP etc..

[Hohn]

A UBB post from Larry Widmer, in case you didn't get to see this when I posted it to the OMG thread... .



Justin







Not having read any of the posted material today, I’ve decided to select a starting place that’s always been a favorite …. Horsepower. What is horsepower other than the formula HP= Torque x RPM / 5252?

Other than something to brag about having, it’s a totally useless term as far as I’m concerned (unless you’re talking steady state, like an airplane, steady cruise, etc. ). I can’t count the number of programs I’ve been involved in that went backwards as we produced more horsepower. Torque is another issue, I suspect you’re thinking about now, and it is, however, perhaps not to the extent you may be thinking.



Any street car or race car is continuously accelerating or decelerating, and, assuming the brakes work and the handling is good, typically, the rate of acceleration is the way we judge the car’s performance and "power". As we’re talking about accelerating a given mass, torque is the most important component, assuming that we’re simply talking HP vs. Torque. If you can’t produce torque at the lower RPM ranges, it’ll take you all day to reach a "higher" RPM where you have torque (or HP. ), and then you shift gears, and you’re back where you started.

If someone is discussing high HP normally aspirated engines of small displacement, I automatically think "RPM" when I hear the numbers, and for those of you unaccustomed to my rambling messages on this board, I do not like RPM. My definition of RPM is Ruins Peoples Motors, and I’ll bet you that you’ll not find any experienced engineer or engine builder who will dispute this notion. I don’t care if you’re discussing imports or domestics, the higher the operating RPM, the higher the stresses, and the sooner you’ll experience parts failures. Before you jump my ass, I think that there are many automobile manufacturers who have done really great jobs in design, material selection, and assembly quality to provide the public with cars with "sufficient" bottom end torque, and reliable upper RPM torque that’s outstanding, and the fact that many modify these engines to produce 2-300% more than the factory anticipated is further testament to the designed and built in quality. However, my point remains relative to RPM. Take two identical engines and run them each "loaded", with one at 4000 rpm and the other at 6000 rpm. Which one would you bet will fail first?



Every program I’ve participated in since 1969 has been designed to build engines that run on torque, and lower rpm than the competition’s engines. If you have torque, you simply gear the car accordingly, and you’ll match the competitions speed, and more importantly, you’ll be more likely to be running at the finish. We used to "recalibrate" the tachometers in our NASCAR, (old) INDY, and ProStock programs to read 1000 rpm high, so when other teams would "casually" look at the tattle tales after running (ProStock was a hoot, as the competition could watch the tach. on the on-board TV broadcasts), and they’d all go deep gear there cars and blow their engines. I damned near shut down some manufacturer’s teams in several racing series because the foundries couldn’t provide enough blocks to replace all the blown ones.



Let’s go to the mystery area that’s really what makes the car run…. or accelerate. I call it Transient Response or Recovery Time. How long does it take for an engine to recover from being yanked down 2000rpm on a shift and accelerate back to the redline again…. that’s the mysterious quality that all killer engines have, and it’s not something that will show up in "conventional" dyno testing. Quite the contrary, I’ve NEVER seen an engine that possessed this quality make big HP. We rarely even look at HP#’s when testing (I stopped that in ’77). The only "number" I’m concerned in is: How much time does it take this engine to accelerate a given "load" from point A to point B. If the rpm range you anticipate operating in is for instance from 5000 to 8000rpm, the combination that will pull the "known load" from the bottom to the top the quickest will ALWAYS be the engine to run regardless of HP. Several years back, one of our NASCAR programs was so dominant that everyone said we had at least 695 HP. to run the way we did, and I will admit that we did have some engines that approached that number, however, all the "big HP. " engines required more than 1. 5 sec. more time to accelerate the load from 6000 to 7600rpm than another "special" engine we’d built. The "special" engine ran almost 10 mph higher lap speeds and it never made more than 590 HP. on the dyno when we decided to "check".



Transient Response is going to become a major topic of discussion in my future writings, as will torque and HP(for what it’s worth). We’ll travel through cylinder heads, manifolds, cams, engine geometry , etc. and we will continuously look at how each affect the "big picture" which is "what happens when you nail it".

Next topic will be more generic: What ingredients are necessary to make "useful" power, and what is "useful" power. …………



The OLD One

 

[Hohn]

How to build a Pro Stock racing engine..

I’ll begin this by first saying that there aren’t many forms of internal combustion engines, or any form of motorsports where the engine development programs have been any more challenging than in NHRA Pro Stock competition. I fully realize that this statement will draw fire, but regardless of the fact that these engines only power a vehicle a quarter mile at a time, and this is an "unlimited heads up" class, the rules governing the configuration of these engines make the design and construction of a competitive power plant very difficult. All the rules governing other "unlimited" type engines are considerably less restrictive in all other forms of competition, so I think the best place to start is with the rules.

A quick run down on NHRA"s Pro Stock engine rules will "read" like this: Engine. 90 degree automotive-type V8, reciprocating, normally aspirated, single distributor, internal combustion engine. Block can be any material or manufacture, single cam. Max. bore spacing of 4. 90", and max. total displacement of 500cid. Two automotive type 4 bbl. carbs, with any internal mods. No split carbs, no fuel injection (yet). Intake manifolds are unlimited in configuration, as long as they’ll fit under a single opening hood scoop with a max. height of 11". Heads must have a max. of two valves, one spark plug, and they must be "castings" with a manufacturer’s logo and pt. #. Now that’s about it, other than the fact that the fuel must be gasoline, and all components such as balancer, and flywheel must be approved. There are also fuel system safety rules, ground clearance, etc.

So, here’s the deal. We need an engine that will propel a 2350lb. , front engine car down a quarter mile in less than 7. 00 sec. at 200 mph. How to do?? First we study the aerodynamics of the car, the max. allowable hood scoop height (dictated by driver vision and aero studies. The chassis configuration is next with finite stress analysis a must to determine flex and resultant suspension reaction abilities. The transmission type is another variable, although, we all are using clutchless 5 speed square cut boxes currently. And then there’s the clutch. I don’t care how much power you can produce, if you can’t make the clutch slide the correct amount on launch, and each gear change, you won’t qualify. Today the 16 qualifiers are typically running within a . 05 sec. separation, and there are 20 more that were within . 10 sec. that will either stay and watch or head for home. Pretty competitive, relative to any other form of motorsports competition, especially when you consider that they don’t have additional laps to make up for driver error.

As I’ve never been high on running excessive rpm, we’re going to attempt to build an engine where I dictate the torque curve. Now this is a different approach for most people, since they usually adapt the car to whatever the engine "dictates", but engines are stupid, and my way is considerably less time consuming and less costly. Typically Pro cars are running almost 10,000 rpm which is really incredible when you consider the displacement and the fact that they are single cam push rod engines. The fact that they’re carbureted has absolutely no effect on their power when compared to any known digital fuel injection, assuming you’ve designed a proper intake manifold, and you understand the "black art" of carburetors.

After measuring both scoop height effectiveness and drag, the real test is the driver and his vision. It’s not likely that you’ll be afforded the luxury of the 11" scoop, but the height we can work with when combined with bottom end clearance and crank C/L will allow a maximum "package height" of 24. 5 (the maximum distance from the top of the intake manifold to the crank C/L) and this number still provides the proper room on top for the carbs. and scoop relationship to work well. Next the decision to run lower rpm is examined, and a torque number is determined to be the minimum to push our combination of frictional losses, drag coefficient and frontal area to speeds of 200+. That torque value at 9,000 rpm is 792 ftlbs. , or 1,360 hp. for those impressed by that #. As we’re going to run less rpm, our rpm range will be slightly wider by necessity, so the math shows that the lowest rpm we’ll encounter is 7,800. The torque necessary to overcome the shift is 852 ft. lbs. which for those who care is 1,265hp. Nasty numbers for a Detroit relic!

The bore / stroke combination for this rev. range will use a slightly longer stroke, and smaller bore, at first the first "look". This is where the package height starts affecting everything we do. Since I attempt to run a rod length to stroke ratio of 1. 75 - 1, calculating the rod length combined with the compression height(distance from pin C/L to deck), and (. 5 x stroke) will yield our block deck height once we settle on the stroke. The combination using 4. 6" bore, and 3. 75" stroke, with a 6. 5" rod, and a piston with a 1. 265" compression height will yield a deck height of 9. 635"and the rod length to stroke ratio will be 1. 73 -1 which is "good". As we’re looking for a really "fat" torque curve, we’re going to build another combination for this engine. Number (2) engine , combined will use a 3. 80" stroke and a 4. 57" bore combined with a 6. 57" rod will again provide a rod to stroke ratio of 1. 73 -1. In order to place each engines internals under the same deck height, the compression height on #2 will be 1. 165" which is "doable".

So "derived from experience" I’m going to build a 498 cid. engine with two different strokes and bores. The two groups of 4 cyls. displacements will be within . 01 ci. coming in at 62. 29 ci. and 62. 28 respectively. The geometry is 1. 73 -1 on each, and the ring package will be placed in the same position despite the different compression heights. Both bores are large enough where the flow rates and swirl characteristics will be very nearly the same.

The bottom end will use a cast iron block which has been "seasoned" and stress relieved. All finish machine work will be done in house including main bearing bores, cam bore (which we’ll raise) lifter bore location, all oil passages, cylinder bores, bolt holes, and all finish interior and exterior machining. The exterior milling is for weight removal and once the mechanical operation is complete, the exterior is chemically treated to remove more material and to dispose of any tool marks which could serve as origins for stress cracks. The cam is moved up to shorten the pushrod length, and the lifter bores locations are also dictated by valve and rocker arm position. We want everything as direct as possible and no push rod angle relative to both lifter body and rocker arm. The camshaft will rotate in needle bearing housings pressed in the block, do oil isn’t a great issue. The lifters are of the roller variety, and their bodies are about twice the diameter of "stock" lifters. The lifter body is made of beryllium for weight purposes, and lifter rotation is prevented by a pin similar to a key which will ride in a slot in the bore.

(continued next post)

 

[Hohn]

The crankshaft is a non twist forging from a Japanese vendor. After rough machining it’s checked for internal and external flaws, and then the finish machining begins. We’ll use the big block Chevy main bearing size, and the rod journal will use a bearing which will have adequate width, but a specific diameter calculated to reduce bearing speed, and resultant losses. Once all the bearing surfaces are "close" the final shape of the counter weights are determined and they are machined to an aerodynamic configuration.

We elect to use titanium rods which will allow us to run tighter clearances than their aluminum counter parts. (piston to head @ . 30" vs. . 90") These rods cost three times more but they’re good for three times the life, or about 45 runs. The entire rotating assembly is internally balanced and ready for assembly.

The oiling system consists of a five stage dry sump pump, with external oil storage tank. The oil pan is designed with a large "kick out" on the rt. side where the oil pick-ups are located. We also fit a scraper to the shape of the rotating assy. That has about . 050" clearance to shear oil from the crank / rods. The bottom of the oil pan is configured and coated to prevent oil from bouncing back up into the rotating assembly. Our pump not only removes oil but one stage is used to create a "vacuum" in the crankcase, the rest is used to pump oil back in. The engine will never have more than 1 quart of oil in it at any given time or rpm.

On top, things get interesting, because now we’re dealing with the piston domes, heads, valve events and manifolding.

During the selection of "brand" we automatically were zoned in on about 4 different heads which are legal. As we’ve had experience with these pieces before and we do not want to do a lot of welding, we’ll make runner, chamber, and water jacket cores which will fit the OE core box and provide heads with adequate material everywhere we need it. The "desirable flow #’s at all valve lifts are calculated, and the port angles and shapes are modeled to verify the characteristics. This is a critical stage as we must also design the intake manifold at the same time, and, remember, it all has to fit in the "package height" which was dictated at the beginning. We begin with the plenum top and overall volume. The volume is dictated by displacement, rpm and rod / stroke ratio, so it’s "fixed", however, a carbureted engine must have the correct runner length, angle, volume, and opening relationship relative to the carbs. throttle bores or the "vacuum signal" will not permit proper carb. function, and the engines ability to accelerate and recover from an instant drop in rpm (on shifts) will be non existent, regardless of torque and power #’s. Now we’ll begin at the top and work down. Once we have the carbs positioned correctly for easy runner entry access, we’ll determine if the intake port previously modeled can coexist with a runner of this angle. If so, we’re OK, but typically we need to go back and redesign the port to compromise its shape to allow it to be properly manifolded. We’ll go round and round and finally settle on a best of the lot. Keep in mind that this is a push rod engine, so the ports must be placed between the push rods, and remember the push rods must be as straight as possible, or valvetrain life is going to be non existent.

Things are now going to become a little more challenging. Remember that the engine has two separate sets of cylinders that do not share the same bore and stroke dimensions. The manifold runners also need to be different lengths and volumes to accomplish what we’re asking. So now we need four runners of one configuration, and four of another. We determine that the short stroke port / runner volume should be 1038. 2 cc’s, and the long stroke will want 1128. 6 cc’s. These volumes should maximize the cylinder’s out put in the desired rpm range.

We now design the combustion chamber and piston dome. We have various models which are known quantities, so we’ll select one that at first pass appears to want to work with bore size, port location (push rod location),etc. The primary goal is to create a really fast burn in a "big" bore, and to also allow great tuning tolerance. The other requirement is that we achieve maximum cylinder pressure well past TDC, at 20 to 30 degrees if possible. To achieve all this we’ll make sure that at TDC there’s no secondary pocketing. The chamber and piston shapes will accelerate the swirl initiated by the intake port’s flow bias, and then concentrate the swirling mixture into a small "sweet spot" near the exhaust valve, and to make things lively we’ll also put the plug in the "sweet spot" as well. The chamber shape now dictates valve location and size, however, if the valve size isn’t sufficient, we must return and look at remodeling the chamber.

I believe we have a chamber and piston dome that’ll work, and based on rpm and relative efficiency, the flow rates are determined for all areas under the lift curve, which is a function of how quickly I can open the intake valve and not exceed piston velocity for the first 12 - 14 degrees of crank rotation. Each intake valve will have it’s own lobe configuration and timing. This may seem acceptable for the differently configured cylinders, but it’s necessary to make changes on every cylinder as well to achieve our power range goals. The valve sizes will be 2. 48" for the intakes and 1. 87" for the exhausts. These aren’t as tight a fit in the bores as you may think, as there’ll be some compound angles involved so the valves will open to the center of the cylinder, and move away from the cylinder wall as they open. The "exit" area size for the exhaust port will be 3. 04 sq. in. , and the area inside the port will be as small as 2. 0 sq. in. , and I can’t tell you where relative to the port length as it’s highly proprietary.

The intake port cross sectional area will range from 4. 84 sq. in. to a minimum of 3. 23 sq. in. , and again, I can’t say where these areas are located in the port. In the intake plenum chamber the runner entry will be 6. 7512 sq, in. with a 1. 46" radius around the runner entry.

There’s also a "black art" that I’ll mention regarding state of the art cylinder heads, and that’s steam manifolding. The design and construction are a pure art form, and if it’s not done correctly, you won’t be in the hunt. That’s all on that subject.

The flow rare for the intakes is 618 cfm @ . 800" lift and 675cfm @ . 800" for the exhaust ports. These values are only for comparison and are at a pressure drop of 28" H2O. We develop and flow components at "different" pressures than the "norm". The exhaust port will reach 85% of the above # by . 400" lift. These lift #’s are not necessarily indicative of the lift #’s currently run in prostock as some cams yield over 1. 0" net lift. I will say this, however, I feel the same way about valve lift as I do about rpm.

The material for intake valves is beryllium and the exhaust valves are titanium. The retainers are titanium, as are the valve springs which need replacement every 12 runs, and you already know the price. The shaft mount rockers are also beryllium, and the pushrods are composites.

The use of ceramic and non friction coatings is extensive on piston domes, skirts, combustion chambers, and valves. We also coat bearings, oil pump components, etc. with moly coatings.

The headers are 2. 5" diameter primary tubes @ 25" long, the collectors are 4. 5" dia. And 13" long after the merge is complete, and all exhaust components are ceramic coated





The carburetors are Holley 4500 series units that flow @ 1300 cfm. each. Fuel systems require a minimum 1" diameter line from the fuel cell to the carb logs, while maintaining constant pressure of 8 psi. The vent on the fuel cell must be at least . 625" or you’ll starve the engine. The ignition is crank triggered MSD with multi- step rev limiters, and timing alterations.

Typical procedure is to break in the rings with a mild cam, "break-in "heads and induction components. Once complete and we see negative crankcase pressure, meaning the rings have seated, it’s time to see if the combination works.

After installing all the "good" stuff, we’ll do a quick run in to re-torque and reset the valves. The next hour is spent installing pressure transducers for each cylinder, exhaust temp. sensors, and chemical analysis probes in each primary tube, correct air ducting and inlet temp. and volume measuring instruments, and also the fuel flow and all normally used sensors. We have the ability to not only adjust carb mixture remotely, but we also can re-map the ignition timing. Dyno’s have come a long way during the last twenty years, and . Dyno’s have come a long way during the last twenty years, and we’ve attempted to stay on top of the game, as the more you can simulate in the dyno cell, the less track time is required. Aside from being able to program a drag strip, oval, or road course into the system, we also simulate G’s by mechanically tilting the entire dyno / engine combination. For lateral G. the engine is all we tilt.

We now have the ability to not only map cylinder pressure from each cylinder vs. crank angle at all rpm. , analyze the combustion efficiency by looking at the gasses that are spent, but we can also read the torque of each cylinder during the tests, and when the test is complete, we can down load, and determine where we need to fine tune.

This particular engine produced slightly better numbers than we "dictated" on it’s first "run". We made some slight mixture and timing changes and found less torque but quicker acceleration. After another hard look at all the numbers, I felt that we needed to alter the cam’s lobes slightly to increase the torque overlap from the different 4 cly. Engines, but also the individual cylinders needed slightly more "spread".

 

[Hohn]

Once installed and all was re-set, we realized considerably more torque than earlier(which was a surprise), but the response time was the best we’ve seen from a Pro engine. Not only did it immediately get up after the simulated shifts, but the transient response (acceleration time from point A to B) was incredible. We also observed that the best timing was 12 degrees BTDC, and the exhaust temps. were down around that 800 degree # we like. Chemically, it was also very clean so we were not only purging the area above the top ring, but also there was little that wasn’t burned. The optimum BFSC’s were all under . 340 which is very good for any engine, and typical of what we’ve been seeing for the last 15 years or so. Most drag racers don’t care much about mileage though.

We did a few tests to determine where the engine became semi-resonant, and although it’s a large engine, the torque extended down to some lower rpm ranges than expected, and the combustion efficiency was surprisingly good in the lower rpm ranges. Although I’ve never been to build two engines that will produce identical power, it was decided that since almost everything was digitized, we’d build a clone to play with while freshening up the customer’s engine. Or is that engine(s)???

And for all those who want compression ratio numbers, the "combustion space" volume for each cylinder equaled 43. 4 cc including the head gasket. If you calculate it, please don’t tell me what it is, as it might worry me and my customers! The gas requirement was 113 octane, and the individual cylinder ignition timing was as low as 11, and as high as 13 degrees…. different engines, every cylinder.

………………………………………. . T. O. O. …………………………………………

 

[esommer2500]

:--) Oo.



Wow. It's going to take me several re-reads and some book time to decipher it ALL, but that was some of the best reading I've had in my college experience!



Eric

 

[Rman]

You guys ever read any of smokey yunick's stuff? He had a lot of "opinionated" stances, was of the old school and knew his stuff. Good readin, thanks hohn. I don't think i've ever met someone so emphatically focused on HP vs. Torque!

 

[snowracer69]

Good reading, and I understand most of it, which makes it even better!



Horsepower sells cars, torque moves mountains and makes trap speeds.



Josh

 

[ThrottleJockey]

Hohn,



So where is the timing set? Would you say a bit less advanced than you originally planned? Or no?



Great reading!

Brian

 

[Hohn]

On the Pro motor larry wrote about above, the timing is varied from cylinder to cylinder. But they are right around 12 degrees BTDC. This is a LOT less timing advance than other gasser performance engines use. Most gassers I have seen use between 28 and 38 degrees of timing lead.



justin

 

[JoshPeters]

I want more! Torque readings? or is it immeasureable? HP just cause it's measureable? I love reading this stuff!





Josh

 

[Shovelhead]

Torque wins races, Horsepower wins Dyno wars.

 

[Steve M]

What I would really like to see on the dyno is usable information. Like how much torque a truck applies to the rear wheels at 70mph. Whether that torque is achieved through gear reduction with a high rpm HEMI or through brute torque like the Cummins makes, how much torque a truck is capable of applying to the rear wheels at a given speed is what's useful. The results might surprise you.

Dyno results are numbers calculated using engine rpm and torque at the wheels. Great for bragging rights but useless otherwise. What is the actual torque at the wheels at what speed?

Theoretically, a HEMI powered 2500 is capable of applying more torque to the road than my SO Cummins powered 2500 at 70mph. I believe a dyno would prove it if it recorded real torque at the wheels. The HEMI can use gear reduction to make torque while my Cummins has to be in O/D at 70mph reducing torque to the drive shaft by 31%.

At 70mph, here are the number using advertised torque.

Cummins SO 4. 10 automatic. 460ft/lbs

460 X . 69 X 4. 10 = 1301 ft/lbs at the rear axle.

HEMI 5. 7 4. 10 automatic. 375ft/lbs at 4200rpm. 325~ at 3k rpm.

325 X 4. 10 = 1332 ft/lbs at the rear axle.

With the HEMI, I could even grab second gear if I needed more torque multiplication. With the Cummins, I can't even drop out of O/D. Drivetrain losses should be similar in both trucks and actual flywheel torque in both is probably less.

Although the HEMI won't last as long or get as good fuel mileage, it has the ability of doing the same work my SO Cummins does by using it's high rpm to make torque through gear reduction.

Just my thoughts while the Bucs play...

 

[Steve Campbell]

That was really.......

"neat"

 

[Hohn]

Originally posted by Steve M

What I would really like to see on the dyno is usable information. Like how much torque a truck applies to the rear wheels at 70mph. Whether that torque is achieved through gear reduction with a high rpm HEMI or through brute torque like the Cummins makes, how much torque a truck is capable of applying to the rear wheels at a given speed is what's useful. The results might surprise you.

Dyno results are numbers calculated using engine rpm and torque at the wheels. Great for bragging rights but useless otherwise. What is the actual torque at the wheels at what speed?

Theoretically, a HEMI powered 2500 is capable of applying more torque to the road than my SO Cummins powered 2500 at 70mph. I believe a dyno would prove it if it recorded real torque at the wheels. The HEMI can use gear reduction to make torque while my Cummins has to be in O/D at 70mph reducing torque to the drive shaft by 31%.

At 70mph, here are the number using advertised torque.

Cummins SO 4. 10 automatic. 460ft/lbs

460 X . 69 X 4. 10 = 1301 ft/lbs at the rear axle.

HEMI 5. 7 4. 10 automatic. 375ft/lbs at 4200rpm. 325~ at 3k rpm.

325 X 4. 10 = 1332 ft/lbs at the rear axle.

With the HEMI, I could even grab second gear if I needed more torque multiplication. With the Cummins, I can't even drop out of O/D. Drivetrain losses should be similar in both trucks and actual flywheel torque in both is probably less.

Although the HEMI won't last as long or get as good fuel mileage, it has the ability of doing the same work my SO Cummins does by using it's high rpm to make torque through gear reduction.

Just my thoughts while the Bucs play...

Steve, great post. You've got me thinking.



The numbers you posted make sense, but it appears that the hemi in your example is in direct, while the CTD is in OD @70 mph. In your example, the advantage of the CTD is in fuel economy and durability. In the example, the Hemi SHOULD equal the towing performance of the CTD, right?



Well, I am not so sure. One thing you have me thinking about is the difference between static and dynamic force. In physics, there is always a differentiation made between static friction and dynamic friction. Both are forces. Static friction is measured by the force required to move a stationary object (accounting for weight or gravity). Dynamic friction is the force required to KEEP that object in motion. This is usually LOWER than the static force.



If i may, I would like to compare this to HP vs tq. I think that the formula of Hp= tq*rpm/5252 does a great deal to confuse people and cause a lot of argument about whether HP or torque is better. Here's an example of how it misleads:



Consider two engines for use in a racecar that we are setting up for a 4000 rpm power band. One engine's powerband is from 2000-6000 rpm. The other engine's powerband is from 5000-9000 rpm. Dyno testing shows BOTH engines to have IDENTICAL and PERFECTLY FLAT torque curves, maintaining an even 400lb-ft across their respective RPM ranges.



So which of the above engines would be faster when installed in the car if all the driveline ratios were the same? Can we even know? What if we installed them in such a way as to where both engines were always within their powerbands?



From a tq perspective, the engines are identical. But because one engine revs a lot higher, it "makes" much more hp, according to the formula. And we COULD gear the car accordingly, and it should be faster, right?



Well, there's no way to know, and that's because of the flaws of the formula for hp in capturing the CONCEPT of power. The formula says that the higher rpm car should be faster, but is it really?



What is power? Simply, power is a RATE. Hp is the RATE of torque (force) application. By definition then, HP is DYNAMIC, or having to do with transitions. But the formula doesn't account for that. If i can produce 100lb-ft at 5252 rpm, then i ALWAYS have 100hp-- whether or not the engine can rev quickly.



I keep coming back to the flywheel example. Obviously, I could fit a very heavy flywheel to an otherwise "powerful" engine. Does the hp of the engine change? NOPE. If I rev it to it's HP peak, it will still "make" the same amount of "power".

But obviously the power DELIVERY of the engine will suffer with the heavy flywheel. This delivery is the RATE of force application.



This is why the formula is wrong. The formula says that the RPM of the engine is the "rate", when it SHOULD be the rate of CHANGE in RPM.



If you are in a car traveling 60mph, it's not the same as if you are accelerating, is it? So we are concerned not with steady state (velocity) but with the transient ( change in velocity)



In physics, there is a sequence of derivatives. Velocity is a rate of change in position. Acceleration is a rate of change of velocity.



So we should only concern ourselves with DYNAMIC forces for torque, not with static.



This goes back to an earlier post I made about DynoJets. While a dynojet might not accurately measure HP, it can tell you how quickly an engine will accelerate a known load, which is MUCH more important than a HP number.



A load dyno like a Mustang will give you an idea of the STATIC torque capability of an engine (the ability to maintain an rpm). This will help you guage towing performance. A dynamic dyno like a dynojet is better for giving you an idea of the ACCLERATION (the ability to raise rpm under load) of an engine.



The two are NOT the same-- and you can't just simply calculate the other because the "formula says so".



Justin

 

[Steve M]

I put the HEMI in direct because it "can" run in direct at 70mph with 4. 10s. My Cummins cannot. It won't turn enough rpm.

To go back a few years, I traded a '94 Ford F-150 with a 300cid I-6 and a 5spd/3. 55 ratio for a '95 Dodge Cummins with auto and 3. 54 ratio. To my disappointment, the Cummns only pulled my travel trailer "as good" as my F-150. The reason being, the Cummins, although it advertised 400ft/lbs of torque, only had 160hp at 2500rpm. I had to run it in O/D even with the 3. 54. So knock 31% off of 400ft/lbs for O/D and I was down to 276ft/lbs. The 300 six was good for 4th gear or even 3rd at 70mph if I needed to downshift for a hill. I could take advantage of it's 265ft/lbs in direct or even use an under-drive gear to multiply it's torque. With the Cummins, I had to slow down to 55mph to even grab direct drive. Going to an under-drive (2nd gear) to multiply torque required slowing down to under 45mph.

Back to the future... My '03 SO Cummins with 4. 10 ratio tows my 10,400 lb 5th wheel 70mph and gets 10. 5mpg doing it so I'm happy with it. My brother in law tows his 5th wheel with his F-250 V-10 auto/3. 73 70mph too and gets 8. 5-9mpg towing. He's happy with his too. Where the difference comes in is, on a 3% grade or more, I have to slow down so I can lock out O/D. So while I'm down to 58mph, he's still running 70mph using gear reduction and high rpm to "make" torque. Of course he's using some fuel but, the point is, he can. At 4200rpm in second gear, he's in his power band and using gear reduction to multiply torque on top of it. I'm at 58mph and my Cummins is out of breath. Things might be different if we we're on a 12% grade and both in 1st gear. Then I would have more torque available from the engine, equal transmission ratios (gear reduction), and more gear reduction (torque multiplication) in the rear end. But I doubt we'll ever be in that situation... Dynamic friction should be about equal with our two rigs.

The two engines in your example open up the same scenario. The 5,000 to 9,000 rpm 400ft/lb motor can use a 2:1 ratio transmission where the 2,000 to 6,000 rpm 400ft/lb motor uses a 1:1 ratio. So the higher rpm motor can make 800ft/lbs of torque at the tail shaft of the transmission to move the vehicle the same speed if the rear end ratio and tire diameter are the same. That's why racing engines are high rpm engines. Imagine the engine that would be required to produce enough torque to overcome the very tall overdrive ratios to achieve 200mph with a 3500lb car at 2500rpm. With high HP (torque at high rpm) you can make torque through gear reduction. On the other hand, you can't make HP with high torque. You can force high speed through overdrive ratios but then you're over stressing your transmission and performance will be ornery and sluggish.

Your flywheel theory is interesting. I can't help but believe a heavy flywheel would be great for towing and damping the vibrations of our Cummins engines. We are already turning a massive amount of rotating and reciprocating mass. But we have only a small damper on the front of the motor to absorb the power pulses and jiggles and flange out back to mount the clutch or TC to. But like you said, a heavier flywheel would slow the rate of change in rpm and feel sluggish.

I like the idea of being able to record the static torque capability of an engine but, I'm more interested in the the static and dynamic torque capabilities of the entire vehicle. If the drive wheels were attached to the flywheel of the motor like on a locomotive, engine HP and torque would be useful. But to read Torque at the rear wheels and engine rpm at the engine only to "calculate" engine HP and torque is totally missing the point for towing applications. And probably not very accurate. How much torque can my truck put to the rear wheels at 70mph would be very useful. My truck as advertised should be able to put 1301 ft/lbs of torque to the pavement. If I put my truck on the dyno, that's what I want to know if it's doing or not.

I get the feeling there's some misunderstanding of what "HP and torque at the rear wheels" means. I asked here a few months ago and have done alot of reading since. I understand gear reduction and torque multiplication as I use these principles in my line of work. So when I started reading of 1,000 ft/lbs of torque "at the rear wheels", I was alittle perplexed. Since it would be very impractical to remove the motor and put it on a dyno stand to read flywheel hp and torque, the dyno uses engine rpm and torque read from the rear wheels to form an equation. The dyno knows what the gear ratios are simply by the speed of the engine verses the speed of the dyno drum and tire diameter. So the HP and torque readings aren't actual, they're calculated.

A motor producing 1,000 ft/lbs of torque at the flywheel in direct gear with a 4. 10 rear ratio would theoretically be making over 4,000 ft/lbs of torque at the drive axle. Put the transmission in 1st gear and WOW! No wonder big chunks of iron shatter like glass and drive-shafts twist like cork-screws.

 

[Hohn]

I'll add another wrinkle to your excellent post above.



Ever notice on a dyno chart how it's possible for torque to DECREASE as rpm goes up, but STILL HAVE MORE HP?



This is because of what you mentioned-- the RPM factor. If the rate that torque decreases is LESS than the rate at which rpm is going up, then you STILL have more hp at the higher RPM even with less torque.



There's a reason that race engines turn so much RPM, and that's because it's much EASIER to increase RPM than it is to increase tq. So we can "make" hp by revving higher and gearing down.



All a gear does is trade torque for time, in one direction or the other. Depending on whether we underdrive or overdrive, we can get whichever we choose.



This brings me to the MAIN reason that I ordered the 6-speed trans. The narrow RPM range of the CTD means that you need as many ratios as possible to best use that rpm range.



I can't say I've ever towed something really heavy. But in my truck (720lb-ft), I have plenty of grunt even in OD. The OD effectively changes the 3. 55 rear into a 2. 56 rear end. That still leaves me with 1843 lb-ft to the rear in OD.



A Hemi with 375 lb-ft would need a 4. 91 final drive ratio to achieve that same torque to the wheels. If the rear end was a 3. 55, then the trans ratio would be 1. 384, which is the same as the NV5600 in 4th gear. This is close to the 1. 45 ratio of second gear in the auto trannys.



So with the same tires and axle ratio, the Hemi needs to turn 3780 rpm (in 4th of an NV5600) or 3972 rpm in 2nd on an auto to equal the performance of a truck like mine in OD at 70mph.



While the Hemi could clearly do this, the Cummins will have a lot easier time of it.



I stick with the diesel... with the stick



Justin

 

[JDMills]

Jeez, I am glued to the computer. Can not add any thing, but having a grand time reading and learning. yahoo! time to revisit the top, and Thank You for sharing.

j

 

[Steve Campbell]

What was the "OMG" thread....

that was mentioned above? It the talk was anything like what I am reading here... . I am hooked